java.security.MessageDigestSpi
java.security.MessageDigest
tribble.security.ZoidbergEHash
|
||||||||||
| PREV CLASS NEXT CLASS | FRAMES NO FRAMES | |||||||||
| SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD | |||||||||
java.lang.Object
Zoidberg-C message digest (cryptographic hash). Computes a 160-bit (20-byte) message digest (a binary hash) from a series of input bytes.
The digest algorithm is designed to produce a unique hash vector for any given input byte stream, such that altering a single bit in the input stream results in a completely different hash value.
An input stream of bytes fed into the digest algorithm are hashed 16 bytes (128 bits) at a time. After all input bytes have been fed into the algorithm, the input is padded with a trailing count byte, and then enough zero bytes to make the total input length a multiple of 16 bytes (128 bits). The result of the digest algorithm is a 20-byte (160-bit) output byte vector.
This digest algorithm can be used as the basis of a digital signature scheme.
Note that none of these methods is synchronized.
Test Vectors
An empty input vector of zero bytes
hashes to a message digest value of
+REDO
{E0 47 A9 DC 13 99 0F A3 06 82 21 D8 7E BA 28 0B 44 17 9A 27}.
A single input byte of
{00}
hashes to a message digest value of
+REDO
{C6 7F 21 C7 7A 7C 4C F1 F5 7F BB 96 B8 21 86 07 85 22 60 5D}.
A single input byte of
{01}
hashes to a message digest value of
+REDO
{35 CB 1D 10 D4 2A 1E 49 43 69 9A DF 9F E4 66 F3 E2 B7 6A 22}.
Two input bytes of
{0D 0A}
hash to a message digest value of
+REDO
{2D 22 DA B7 0C 7F C4 D7 E5 13 57 93 7C DD 4B 21 1C B1 24 2F}.
Four input bytes of
{01 02 03 04}
hash to a message digest value of
+REDO
{76 C9 35 85 86 08 70 86 47 39 5C FE 4C 24 0C CC 4F BE 80 61}.
16 input bytes of
{01 02 03 04 01 02 03 04 01 02 03 04 01 02 03 04}
hash to a message digest value of
+REDO
{B4 97 4B 0D 4B FF 28 94 C2 AA C8 43 F2 2B EC 4F 3E 45 25 43}.
16 input bytes of
{00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F}
hash to a message digest value of
+REDO
{EB 6D 0C 0C 8E C5 B5 44 48 57 EC 73 63 3D 1E 1C 0E E8 30 26}.
16 input bytes of
{00 01 02 03 04 25 06 07 08 09 0A 0B 0C 0D 0E 0F}
hash to a message digest value of
+REDO
{7E 03 5F 44 C8 F7 5A 07 E3 B5 08 F7 3D 50 1C EC D1 6B 0D EF}.
17 input bytes of
{00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 00}
hash to a message digest value of
+REDO
{0D 90 9E CF F4 15 14 89 89 95 6D 0D 06 40 33 5E 8D 45 B1 B9}.
16 input bytes of
{44 61 76 69 64 20 52 2E 20 54 72 69 62 62 6C 65}
hash to a message digest value of
+REDO
{76 8B 2B 77 37 53 72 89 7F 78 14 8F A3 57 12 CA 18 D7 AE 9A}.
Acknowledgments
This code is based on an original algorithm invented by David R. Tribble in March 2002.
License is granted to use this algorithm without fees or restrictions for all private and commercial uses.
ZoidbergAHash,
MessageDigest| Constructor Summary | |
ZoidbergEHash()
Constructor. |
|
| Method Summary | |
java.lang.Object |
clone()
Clone this message digest, making a copy of the contents accumulated up to this point. |
byte[] |
digest()
Compute the final value of this message digest. |
byte[] |
digest(byte[] in)
Hash an array of bytes, updating and computing the final message digest. |
protected byte[] |
engineDigest()
Compute the final value of this message digest. |
protected int |
engineGetDigestLength()
Retrieve the size, in bytes, of this message digest. |
protected void |
engineReset()
Reset this message digest, clearing the results of any previously accumulated hashed input data. |
protected void |
engineUpdate(byte in)
Hash a single byte, updating this message digest. |
protected void |
engineUpdate(byte[] in,
int off,
int len)
Hash an array of bytes, updating this message digest. |
protected byte[] |
getBytes()
Retrieve the output bytes of this message digest. |
protected int |
hashWords()
Hash four 32-bit input words (128 bits), updating this message digest. |
static void |
main(java.lang.String[] args)
Test driver. |
void |
reset()
Reset this message digest, clearing the results of any previously accumulated hashed input data. |
void |
update(byte in)
Hash a single byte, updating this message digest. |
void |
update(byte[] in)
Hash an array of bytes, updating this message digest. |
void |
update(byte[] in,
int off,
int len)
Hash an array of bytes, updating this message digest. |
| Methods inherited from class java.security.MessageDigest |
digest, getAlgorithm, getDigestLength, getInstance, getInstance, getInstance, getProvider, isEqual, toString |
| Methods inherited from class java.security.MessageDigestSpi |
engineDigest |
| Methods inherited from class java.lang.Object |
equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait |
| Constructor Detail |
public ZoidbergEHash()
| Method Detail |
public static void main(java.lang.String[] args)
Reads bytes from the standard input, hashes them, and then prints the resulting 160-bit (20-byte) hexadecimal hash vector.
args - Command line arguments. If specified, a message digest is computed from
the first argument; otherwise the digest is computed from the bytes read
from standard input.public void reset()
public java.lang.Object clone()
A cloned copy of a message digest can be used to produce a partial message
digest. Simply clone the message digest after the partial message
contents have been fed into update(byte[]), and then call
digest() on the clone. The original message digest object can
then be used to continue hashing the rest of the message.
public void update(byte in)
in - A single byte of input data to hash.public void update(byte[] in)
in - An array of data bytes to hash.
public void update(byte[] in,
int off,
int len)
in - An array of data bytes to hash.off - The index of the first byte of array in to hash.len - The number of bytes in array in to hash.public byte[] digest()
Note that this also resets the hash vector of this message digest.
public byte[] digest(byte[] in)
Note that this also resets the hash vector of this message digest.
in - An array of data bytes to hash.
protected void engineReset()
protected int engineGetDigestLength()
protected void engineUpdate(byte in)
in - A single byte of input data to hash.
protected void engineUpdate(byte[] in,
int off,
int len)
in - An array of data bytes to hash.off - The index of the first byte of array in to hash.len - The number of bytes in array in to hash.protected byte[] engineDigest()
The input is padded with a trailing count byte, and then enough zero bytes to make the total input length a multiple of 20 bytes (160 bits).
Note that this also resets the hash vector of this message digest.
protected byte[] getBytes()
Note that the returned byte array will contain all zeros (0x00) if this message digest has not been updated with any data bytes or has been reset.
protected int hashWords()
Algorithm
The hashing algorithm is a "braiding" of five 32-bit binary words, which comprise a total of 160 bits of the hash vector. The "braiding" refers to the way in which the words are permuted (rearranged). Between each permuting operation (Pi-phase), the input vector is combined with the output of the permutation, using exclusive-or (X-phase) and two's-complement addition (A-phase).
The I-phase spreads four 32-bit input words across the current 160-bit hash vector. The four 32-bit input words comprise the first four words of the input vector; the last word of the input vector is the sum of the squares of the four input words minus one (i.e., i4 = in0**2 + in1**2 + in2**2 + in3**2 - 1).
I-phase:
+-----+ +-----+ +-----+ +-----+
| in0 | | in1 | | in2 | | in3 | Input words (128 bits)
+-----+ +-----+ +-----+ +-----+
: : : :
: : : : : : : :
: : : : : : : :
: X**2 : X**2 : X**2 : X**2 -1
: : : : : : : : :
: :.... : ..ADD... : ..ADD... : ..ADD....ADD
: : : : :
: : : : :
+-----+ +-----+ +-----+ +-----+ +-----+
| i0 | | i1 | | i2 | | i3 | | i4 | I (160 bits)
+-----+ +-----+ +-----+ +-----+ +-----+
The H-phase adds the input vector words with the previous hash vector.
H-phase:
+-----+ +-----+ +-----+ +-----+ +-----+
| i0 | | i1 | | i2 | | i3 | | i4 | I0 (160 bits)
+-----+ +-----+ +-----+ +-----+ +-----+
: : : : :
: +-----+ : +-----+ : +-----+ : +-----+ : +-----+
: | h0 | : | h1 | : | h2 | : | h3 | : | h4 | Hi-1
: +-----+ : +-----+ : +-----+ : +-----+ : +-----+
: : : : : : : : : :
: : : : : : : : : :
ADD ADD ADD ADD ADD
: : : : :
+-----+ +-----+ +-----+ +-----+ +-----+
| i0 | | i1 | | i2 | | i3 | | i4 | I1
+-----+ +-----+ +-----+ +-----+ +-----+
The R, P, X, Q, and A phases are then repeated 5 times, intermixing the contents of the input vector with the hash vector. The final result of the five rounds is the output hash vector.
R-phase:
+-----+ +-----+ +-----+ +-----+ +-----+
| i0 | | i1 | | i2 | | i3 | | i4 | Ii
+-----+ +-----+ +-----+ +-----+ +-----+
: : : : :
ROR(17) ROR(5) ROR(25) ROR(13) ROR(29)
: : : : :
+-----+ +-----+ +-----+ +-----+ +-----+
| r0 | | r1 | | r2 | | r3 | | r4 | Ri
+-----+ +-----+ +-----+ +-----+ +-----+
P-phase:
+-----+ +-----+ +-----+ +-----+ +-----+
| r0 | | r1 | | r2 | | r3 | | r4 | Ri
+-----+ +-----+ +-----+ +-----+ +-----+
: : : : :
: : : : :
: : : : :
: : : : :
: : :
: : : : :
: : : : :
: : : : :
: : : : :
+-----+ +-----+ +-----+ +-----+ +-----+
| p0 | | p1 | | p2 | | p3 | | p4 | Pi
+-----+ +-----+ +-----+ +-----+ +-----+
X-phase:
+-----+ +-----+ +-----+ +-----+ +-----+
| p0 | | p1 | | p2 | | p3 | | p4 | Pi
+-----+ +-----+ +-----+ +-----+ +-----+
: : : : :
+-----+ : +-----+ : +-----+ : +-----+ : +-----+ :
| i0 | : | i1 | : | i2 | : | i3 | : | i4 | : Ii
+-----+ : +-----+ : +-----+ : +-----+ : +-----+ :
: : : : : : : : : :
: NEG : : : : : : : :
: : : : : : : : : :
: : : : : : : : : :
XOR XOR XOR XOR XOR
: : : : :
+-----+ +-----+ +-----+ +-----+ +-----+
| x0 | | x1 | | x2 | | x3 | | x4 | Xi
+-----+ +-----+ +-----+ +-----+ +-----+
Q-phase:
+-----+ +-----+ +-----+ +-----+ +-----+
| x0 | | x1 | | x2 | | x3 | | x4 | Xi
+-----+ +-----+ +-----+ +-----+ +-----+
: : : : :
: : : : :
: : : : :
: : : : :
: : :
: : : : :
: : : : :
: : : : :
: : : : :
+-----+ +-----+ +-----+ +-----+ +-----+
| q0 | | q1 | | q2 | | q3 | | q4 | Qi
+-----+ +-----+ +-----+ +-----+ +-----+
A-phase:
+-----+ +-----+ +-----+ +-----+ +-----+
| q0 | | q1 | | q2 | | q3 | | q4 | Qi
+-----+ +-----+ +-----+ +-----+ +-----+
: : : : :
+-----+ : +-----+ : +-----+ : +-----+ : +-----+ :
| i0 | : | i1 | : | i2 | : | i3 | : | i4 | : Ii
+-----+ : +-----+ : +-----+ : +-----+ : +-----+ :
: : : : : : : : : :
: : : : : : : : : :
ADD ADD ADD ADD ADD
: : : : :
+-----+ +-----+ +-----+ +-----+ +-----+
| i0 | | i1 | | i2 | | i3 | | i4 | Ii+1
+-----+ +-----+ +-----+ +-----+ +-----+
The output vector of each round becomes the input vector of the next round.
The output vector of the last round is the output of the hash, i.e., the message digest.
+-----+ +-----+ +-----+ +-----+ +-----+
| i0 | | i1 | | i2 | | i3 | | i4 | I5
+-----+ +-----+ +-----+ +-----+ +-----+
: : : : :
+-----+ +-----+ +-----+ +-----+ +-----+
| h0 | | h1 | | h2 | | h3 | | h4 | Hi+1, O
+-----+ +-----+ +-----+ +-----+ +-----+
The P and Q phases permute the 5 words of the hash vector in the following braiding pattern (which repeats every 5 iterations):
0. A B C D E
1. B A D C E P iteration 1
2. B D A E C Q
3. D B E A C P iteration 2
4. D E B C A Q
5. E D C B A P iteration 3
6. E C D A B Q
7. C E A D B P iteration 4
8. C A E B D Q
9. A C B E D P iteration 5
10. A B C D E Q
|
||||||||||
| PREV CLASS NEXT CLASS | FRAMES NO FRAMES | |||||||||
| SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD | |||||||||